The reconstruction of images of an object of interest may introduce artifacts along lines of high gradients of absorption values. According to an exemplary embodiment of the present invention, these artifacts may efficiently be removed by a statistical weighing during reconstruction of the image. Advantageously, according to an aspect of the present invention, the reconstruction of the image may be performed iteratively, wherein the updates are weighted with the intrinsic statistical error of the measured photon counts. This may lead to an efficient removal of artifacts.
|
9. A data processing device, comprising: a memory for storing a data set of an object of interest; a data processor for performing artifact correction in the data set of the object of interest, wherein the data processor is adapted for performing the following operation: loading the data set; reconstructing an image of the object of interest on the basis of the data set; wherein a statistical weighing is performed during reconstruction of the image; wherein the weighing comprises an intrinsic statistical error σY, based on measured photon counts Yi, where σY, is the square root of Yi.
1. A method of artifact correction in a data set of an object of interest, the method comprising the step of: reconstructing an image of the object of interest on the basis of the data set; wherein a statistical weighing is performed during reconstruction of the image, wherein the reconstruction of the image is performed on the basis of an iterative algorithm comprising a plurality of update steps until an end criterion has been fulfilled, wherein each update step comprises subtractions weighted with an intrinsic statistical error σY, based on measured photon counts Yi, wherein σY, is the square root of Yi.
11. An apparatus for performing artifact correction in a data set of an object of interest, comprising:
a processor; and
a computer readable storage medium encoded with computer executable instructions which, when executed by the processor, causes the processor to perform the following operation loading the data set; and
reconstructing an image of the object of interest on the basis of the data set; wherein a statistical weighing is performed during reconstruction of the image; wherein the weighing comprises an intrinsic statistical error σY, based on measured photon counts Yi, where σY,is the square root of Yi.
2. The method according to
3. The method according to
4. The method according to
determining a number of detected photons during acquisition of the data set; wherein the weighing is based on a statistical error of the number of detected photons.
5. The method according to
6. The method according to
7. The method according to
determining a number of detected photons Yi during acquisition of the data set; wherein the weighing is based on a statistical error σy
wherein di is a number of photons emitted by the source of radiation;
wherein lij is a basis function of an i-th projection;
wherein li is a vector of basis functions lij of the i-th projection; and
wherein <li, μ>=Σilijμj is an inner product.
8. The method of
10. The data processing device according to
12. The apparatus of
13. The apparatus of
14. the apparatus of
15. The apparatus according to
16. the apparatus of
17. The apparatus of
wherein di is a number of photons emitted by the source of radiation; wherein lij is a basis function of an i-th projection; wherein li is a vector of basis functions lij of the i-th projection;
and wherein <li, μ>=Σjlijμj is an inner product.
18. The apparatus according to
|
The present invention relates to the field of image processing, for example in medical applications. In particular, the present invention relates to a method of artifact correction in a data set of an object of interest, to data processing devices and to respective computer programs.
In a CT scanner with a polychromatic source of radiation, such as a polychromatic x-ray source, a polychromatic x-ray beam passes through matter and low-energy photons are absorbed as the linear attenuation coefficient generally decreases with energy. As a result, the beam gradually becomes harder, i.e. its mean energy increases. The harder the beam, the less it is further attenuated. Therefore, the total attenuation is no longer a linear function of the thickness of the absorber. This non-linear behavior generally leads to a reduction in the attenuation coefficient. In bone for example, beam-hardening can cause reductions of up to 10%. Neglecting this effect in the reconstruction process leads to well-known beam-hardening artifacts, such as cupping, streaks and flairs.
The reconstruction of CT images can be done with iterative algorithms. These algorithms, in particular iterative maximum-likelihood (ML) reconstruction algorithms, introduce artifacts along lines of high gradients of absorption values. This is well-known from the expectation maximation maximum-likelihood algorithm (EM-ML algorithm). Theses artifacts however are not desired, since they lead to a reduction of image quality.
It is an object of the present invention to provide for an improved image quality.
According to an exemplary embodiment of the present invention, the above object may be solved by a method of artifact correction in a data set of an object of interest, wherein an image of the object of interest is reconstructed on the basis of the data set and wherein a statistical weighing is performed during reconstruction of the image.
Advantageously, this may allow for a correction of artifacts in an image on the basis of a statistical weighing during reconstruction of the image. By weighting weighing certain data in the data set statistically the image quality may be improved significantly.
According to another exemplary embodiment of the present invention, the data set is a projection data set acquired by means of a source of electromagnetic radiation generating a beam and by means of a radiation detector detecting the beam.
Advantageously, this may allow for an artifact correction of CT images acquired in a CT system, which may be used, for example, in the field of medical imaging, and therefore for an improved image quality.
In another exemplary embodiment of the present invention, the source of electromagnetic radiation is a polychromatic x-ray source. Furthermore, according to an aspect of this exemplary embodiment of the present invention, the source moves along a helical path around the object of interest and the beam has one of a cone beam geometry and a fan beam geometry.
The application of polychromatic x-rays may be advantageous, since polychromatic x-rays are easy to generate and provide for a good image resolution. Furthermore, since the geometry of the CT scanner system may be of different designs, such as, for example, cone beam or fan beam geometry, a method of an exemplary embodiment of the present invention may be applied to a plurality of different scanner systems and may not be limited to pure CT scanner systems, but also applicable to positron emission tomography (PET) or single photon emission computed tomography (SPECT).
According to another exemplary embodiment of the present invention, the reconstruction of the image is performed on the basis of an iterative algorithm comprising a plurality of update steps until an end criterion has been fulfilled.
Advantageously, by performing the reconstruction of the image iteratively, mathematical reconstruction methods may be used which comprise transcendental equations which can not be solved exactly. Furthermore, an iterative approach may provide for improved image quality, since the iterative loop may be repeated a plurality of times until the end criterion is met.
According to another exemplary embodiment of the present invention, the iterative algorithm is a maximum-likelihood algorithm and the reconstructed image has the highest likelihood. Furthermore, the weighing is performed in each update step of the plurality of update steps.
Advantageously, a maximum-likelihood algorithm allows for determination of an equation which defines an image with the highest likelihood and a weighing in each update step of the plurality of update steps allows for an efficient compensation or correction of artifacts.
In another exemplary embodiment of the present invention, a number of detected photons during acquisition of the data set is determined and the weighing is based on a statistical error of the number of detected photons.
Thus, the statistical error of the photon counts is estimated and taken into account during the image reconstruction, which may result in an improved image quality with less artifacts.
According to another exemplary embodiment of the present invention, an update of an attenuation parameter μjn+1 is calculated from the attenuation parameter μjn by
wherein di is a number of photons emitted by the source of radiation, lij is a basis function of an i-th projection, li is a vector of basis functions lij of the i-the projection and <li,μ> is an inner product, defined by <li,μ>=Σjlijμj.
This iterative statistical uncertainty weighted algorithm for maximum-likelihood transmission tomography may allow for an effective removal of artifacts by weighing the update with the intrinsic statistical error of the measured photon counts.
According to another exemplary embodiment of the present invention, the reconstruction of the image is based on a sub-set of at least two projections of all acquired projections of the projection data set.
Advantageously, according to this exemplary embodiment of the present invention, performing the statistical weighing on sub-sets of projections may lead to a fast convergence of the image with less artifacts.
According to another exemplary embodiment of the present invention, a data processing device is provided, which comprises a memory for storing a data set and a data processor for performing artifact correction in the data set of the object of interest, wherein the data processor is adapted for performing the following operation: loading the data set and reconstructing an image of the object of interest on the basis of the data set, wherein a weighing is performed during reconstruction of the image and wherein the weighing is based on statistical considerations.
Advantageously, this may allow for an improved image quality of images processed by the data processor which comprise artifacts caused by, for example, high attenuating objects, such as metal.
According to another exemplary embodiment of the present invention, the reconstruction of the image is performed on the basis of an iterative algorithm comprising a plurality of update steps until an end criterion has been fulfilled, wherein the iterative algorithm is a maximum-likelihood algorithm and the reconstructed image has the highest likelihood. Furthermore, the weighing is performed in each update step of the plurality of update steps.
Advantageously, the iterative approach may allow for an improved and fast convergence of the image with less artifacts.
According to another exemplary embodiment of the present invention, a CT scanner system is provided, comprising a memory for storing a data set of an object of interest and a data processor for performing artifact correction in the data set of the object of interest, wherein the data processor is adapted for performing the following operation: loading the data set and reconstructing an image of the object of interest on the basis of the data set, wherein a statistical weighing is performed during reconstruction of the image.
Advantageously, this may allow for an improved CT image quality of for example a part of the body of a patient, wherein the acquired data is processed by the data processor which comprise artifacts caused by, for example, high attenuating objects, such as a prosthesis or other implants.
The present invention relates also to a computer program which may, for example, be executed on a processor, such as an image processor. Such a computer program may be part of, for example, a CT scanner system. The computer program may be preferably loaded into working memories of data processors. The data processors are thus equipped to carry out exemplary embodiments of the methods of the present invention. The computer program may be written in any suitable programming language, for example in C++and may be stored on a computer readable medium, such as a CD-ROM. Also, these computer programs may be available from a network, such as the WorldWideWeb, from which they may be downloaded into image processing units or processors, or any suitable computers.
It may be seen as the gist of an exemplary embodiment of the present invention that artifacts in an image are corrected on the basis of a statistical weighing during the reconstruction of the image. This may lead to an improved image quality with less artifacts.
These and other aspects of the present invention will become apparent from and elucidated with reference to the embodiments described hereinafter.
Exemplary embodiment of the present invention will be described in the following, with reference to the following drawings:
The scanner depicted in
Reference numeral 5 designates an aperture system which forms the radiation beam emitted from the radiation source to a cone-shaped radiation beam 6. The cone beam 6 is directed such that it penetrates an object of interest 7 arranged in the center of the gantry 1, i.e. in an examination region of the CT scanner, and impinges onto the detector 8. As may be taken from
During a scan of the object of interest 7, the source of radiation 4, the aperture system 5 and the detector 8 are rotated along the gantry 1 in the direction indicated by arrow 16. For rotation of the gantry 1 with the source of radiation 4, the aperture system 5 and the detector 8, the motor 3 is connected to a motor control unit 17, which is connected to a calculation unit 18.
In
The detector 8 is connected to calculation unit 18. The calculation unit 18 receives the detection result, i.e. the read-outs from the detector elements of the detector 8 and determines a scanning result on the basis of these read-outs. The detector elements of the detector 8 may be adapted to measure the attenuation caused to the cone beam 6 by the object of interest 7. Furthermore, the calculation unit 18 communicates with the motor control unit 17 in order to coordinate the movement of the gantry 1 with motor 3 and 20 are with the conveyor belt 19.
The calculation unit 18 may be adapted for reconstructing an image from read-outs of the detector 8. The image generated by the calculation unit 18 may be output to a display (not shown in
The calculation unit which may be realized by a data processor may also be adapted to perform an artifact correction in the image based on read-outs from the detector elements of the detector 8. According to an aspect of the present invention, this correction may be performed by reconstructing an image of the object of interest on the basis of the read-outs, wherein a statistical weighing is performed during reconstruction of the image.
Furthermore, the calculation unit for the data processing device may be adapted for performing artifact correction in the data set by performing the reconstruction of the image on the basis of an iterative algorithm comprising a plurality of update steps until an end criterion has been fulfilled, wherein the iterative algorithm is a maximum-likelihood algorithm and wherein the reconstructed image has the highest likelihood. Furthermore, the weighing is performed in each update step of the plurality of update steps.
Furthermore, as may be taken from
A standard approach for a reconstruction of CT images are iterative algorithms, for example a convex maximum likelihood algorithm. The Maximum Likelihood Method is able to determine an equation which defines an image with the highest likelihood L(μ) by setting ∂L(μj)/∂μj =0, where μ is the vector of attenuation parameters μj. Since this equation is transcendental, it cannot be solved exactly, but an iterative approach can lead to the solution. One of these iterative approaches is the “Convex Algorithm” described in K. Lange and J. A. Fessler, “Globally convergent algorithms for maximum a posteriori transmission tomography”, IEEE Trans. Med. Imaging 4, 1430-1450 (1995), which is hereby incorporated by reference. The update for one iterative step is found to be
where di and Yi are the emitted and observed number of photon counts, respectively, li is the vector of basis functions lij of the ith projection, and <li,μ>is the inner product Σilijμj. This update ensures an increasing likelihood in the neighborhood of the optimal point, thus it should lead to an optimal image.
The reconstruction result of such an iterative algorithm is shown as image slice 31 in
The artifacts in the reconstructed image slice 31 occur as streaks. As may be seen from image slice 31 in
On the other hand, by reconstructing the image slice according to an exemplary embodiment of the method according to the present invention, in which the subtractions in the updates are weighted with the intrinsic statistical error σY
where σY
It should be noted, that the method according to an exemplary embodiment of the present invention also works for a reconstruction with sub-sets of all measured projections.
The image slices shown in
The convex maximum likelihood algorithm used for reconstructing the image slices of
In step S6 then, a counter is set to n=0 after which the iterative loop for the image reconstruction starts. In step S7 the n-th update (which is in case of the beginning of the iterative loop the first update) of the attenuation parameter μj is calculated by
Then, the method continues to step S8, where it is determined whether an end criterion is met or not. The end criterion may, for example, be that the iteration has been performed for a certain predetermined number of times or that a difference between consecutive updates does not exceed a certain predetermined threshold value. In case that it is determined in step S8 that the end criterion is not met, the counter is increased by one (step S9) and the method continues with step S7, where (n+1)-th update of the attenuation parameter μj is calculated. Step S7-S9 may be iteratively repeated until the end criterion is met. In case it is determined in step S8, that the end criterion is met, the method continues to step S10, where it ends.
It should be noted that, as may be apparent to the skilled person, the above described technique may be applied to all known reconstruction techniques in which a difference is summarized such as in the convex algorithm described above.
Furthermore, via the bus system 153, it is also possible to connect the image processing and control processor 151 to, for example a motion monitor, which monitors a motion of the object of interest. In case, for example, a lung of a patient is imaged, the motion sensor may be an exhalation sensor. In case the heart is imaged, the motion sensor may be an electrocardiogram (ECG).
Patent | Priority | Assignee | Title |
10019811, | Dec 22 2014 | Canon Kabushiki Kaisha | Image processing apparatus and method for estimating an error in registration between images |
10939887, | Jun 23 2014 | PaloDEx Group OY | System and method of artifact correction in 3D imaging |
8571291, | Oct 19 2011 | Toshiba Medical Systems Corporation | Combination weight applied to iterative reconstruction in image reconstruction |
8705831, | Sep 24 2009 | Koninklijke Philips Electronics N V | System and method for generating an image of a region of interest |
9364192, | Mar 15 2013 | Siemens Medical Solutions USA, Inc. | Error estimates in quantitative functional imaging |
9418453, | Jul 13 2011 | MILTENYI BIOTEC B V & CO KG | Transmission image reconstruction and imaging using poissonian detector data |
9466135, | Aug 20 2009 | Koninklijke Philips Electronics N V | Reconstruction of a region-of-interest image |
9592020, | Jun 23 2014 | PaloDEx Group OY | System and method of artifact correction in 3D imaging |
Patent | Priority | Assignee | Title |
4149081, | Nov 29 1976 | Varian Associates, Inc. | Removal of spectral artifacts and utilization of spectral effects in computerized tomography |
5416815, | Jul 02 1993 | General Electric Company | Adaptive filter for reducing streaking artifacts in x-ray tomographic images |
6507633, | Feb 15 2001 | The Regents of the University of Michigan | Method for statistically reconstructing a polyenergetic X-ray computed tomography image and image reconstructor apparatus utilizing the method |
6744848, | Feb 11 2000 | Brandeis University | Method and system for low-dose three-dimensional imaging of a scene |
20020154735, | |||
20030128801, | |||
20030156684, | |||
20030219152, | |||
20070019782, | |||
WO2067201, |
Executed on | Assignor | Assignee | Conveyance | Frame | Reel | Doc |
Mar 08 2005 | Koninklijke Philips Electronics N.V. | (assignment on the face of the patent) | / | |||
Aug 01 2005 | ZIEGLER, ANDY | Koninklijke Philips Electronics N V | ASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS | 018232 | /0248 |
Date | Maintenance Fee Events |
Oct 22 2012 | M1551: Payment of Maintenance Fee, 4th Year, Large Entity. |
Oct 24 2016 | M1552: Payment of Maintenance Fee, 8th Year, Large Entity. |
Dec 14 2020 | REM: Maintenance Fee Reminder Mailed. |
Dec 17 2020 | M1553: Payment of Maintenance Fee, 12th Year, Large Entity. |
Dec 17 2020 | M1556: 11.5 yr surcharge- late pmt w/in 6 mo, Large Entity. |
Date | Maintenance Schedule |
Apr 28 2012 | 4 years fee payment window open |
Oct 28 2012 | 6 months grace period start (w surcharge) |
Apr 28 2013 | patent expiry (for year 4) |
Apr 28 2015 | 2 years to revive unintentionally abandoned end. (for year 4) |
Apr 28 2016 | 8 years fee payment window open |
Oct 28 2016 | 6 months grace period start (w surcharge) |
Apr 28 2017 | patent expiry (for year 8) |
Apr 28 2019 | 2 years to revive unintentionally abandoned end. (for year 8) |
Apr 28 2020 | 12 years fee payment window open |
Oct 28 2020 | 6 months grace period start (w surcharge) |
Apr 28 2021 | patent expiry (for year 12) |
Apr 28 2023 | 2 years to revive unintentionally abandoned end. (for year 12) |